 Ash melting temperature is an important parameter for the safe operation of combustion systems. It is typically measured using the STN-ISO540 method in a laboratory. However, this method is costly and time-consuming. A new method based on multiple linear regression has been developed to predict ash melting temperature from the known chemical composition of fuel ash and the amount of SiO2, chow, K2O, MGO and Al2O3 added as an additive. This method was tested against real ash melting temperatures and achieved an accuracy of 90%. This article was authored by Michael Holupsik and Josef Jandaka.